a form of witnessing that contends with and represents difficult questions that are hidden or forgotten. Acknowledging the importance of my experience as meaningful data, does not resolve the ethical decisions and moral choices I made. These ethical considerations were continually renegotiated and lived beyond the life of my NGO work. As this piece of ‘research’ was not approved by any ethical review board, I was left with quotes I did not have the consent to use. Does this mean, then, that using them is inevitably unethical? Could I speak about my perceptions of actors, who did not know that they were research participants? While a positivist view of research would easily dismiss this as non-research, this is an ongoing ethical process for I have wrestled with the dilemmas and contradictions that emerge from researching in this manner. I decided not to name the NGO that I worked at because my purpose here is not to target or expose the action of the employees or the leadership of the organisation. Instead through referring to a range of experiences, including mine, I aim to problematize first and foremost the role of non-state actors in a detention setting. Furthermore, I removed all the identifiers from people I interacted with and I quote as part of this. I have also sought to refer only to events that I was a witness of and not others that were related to me either by detainees and staff or colleagues. To take this further, in order to triangulate events, I wanted to include in my accounts, I actively tried to interview the people, who were involved or were there for me. In all cases but one, I succeeded. To do research where (former) colleagues are participants is to pay attention to the complex power relations inherent is such relationships. Chapter 5, which heavily draws on these experiences, is not simply a critique of others. It places myself at the centre of the discussion and in no way do I hope to evade the criticism I extend to others. By bringing together the academic, the professional and the personal, I seek to make the project ethically reflexive. Navigating between multiple roles can be challenging both for a research design but also for the analysis, highlighting issues of trust and conflicts that may arise between researcher ‘roles’ and practitioner ‘responsibilities’ (Stacey, 1988; Gorman, 2007). As others have shown (Drake and Heath, 2010), as the analysis progressed, my different roles merged so that both research and practice informed each other in the pursuit of new knowledge. 61